STATE OF THE UNION poverty head 15 The Stanford Center on and Inequality BY JANET C. GORNICK AND MARKUS JÄNTTI KEY FINDINGS Using a relative poverty standard for disposable household income, the U.S. poverty rate exceeds that reported in all of the other high-income countries in this study, with the sole exception of Israel. The well-known exceptionalism of American relative poverty extends only to rich countries. Most of the middle-income countries in this study report higher relative poverty rates than are seen in the United States. U.S. children are 30 percent more likely to live in relative poverty than is the U.S. population overall. This general pattern is not unusual. In about threequarters of the rich countries included in this study, children s poverty risk (visà-vis disposable income) is higher than that of all persons. When we consider absolute poverty (using a poverty line based on the official U.S. threshold), American children are more likely to be poor than children in 11 of the 20 study countries. And nine of these 11 countries all but Luxembourg and Norway are less affluent than the U.S. It is well-known, at least among scholars of poverty, that the U.S. has more poverty than most other high-income countries. That result has now been established in a substantial research literature based on data from LIS (formerly known as the Luxembourg Income Study), from the Organization for Economic Cooperation and Development (OECD), and from Eurostat (the statistical office of the European Union). 1 In this report, we take a closer look at that claim, drilling down and extending it in several ways. We first examine just how robust the claim is. We explore whether the U.S. still stands out as a high-poverty country (a) when a broader range of countries, even middle-income ones, are considered; (b) when poverty is defined in terms of both market income and disposable income; 2 and (c) when poverty is defined in both relative and absolute terms. Second, we examine whether the U.S. has a distinctive system when it comes to poverty among children, a subgroup that draws our attention due to the worrisome consequences, both short- and long-term, associated with child poverty. We know that U.S. children are 30 percent more likely to live in poverty than is the U.S. population overall. Here, we ask: Is that a widespread pattern? In other words, is poverty among children disproportionately high within other countries too? We next assess the association between poverty risk and household structure. The key question here is whether single mothers in all countries, not just the U.S., are exposed to a disproportionate risk of poverty. We thus start with the finding that, in the U.S., households headed by single mothers are nearly four times as likely to be poor as are those headed by two parents. We again ask: Is that a widespread pattern? The risk of poverty in the U.S. is also strongly affected by educational attainment 3 and attachment to the labor market. 4 In the U.S., persons without high-school degrees are more than six times as likely to be poor as those who have completed post-secondary school, and persons with no earnings are 80 times as likely to live in poor households as are those whose earnings place them in the top two-thirds of their earnings distribution. 5 Again, we ask: Are these disparities including the steepness of the gradients widespread? Brief Remarks about Our Empirical Work All of our results are based on microdata contained in the LIS database. 6 The first table, which includes 34 countries, pertains to households with heads of all ages. In subsequent tables (Tables 2 5), we restrict our analyses to younger households and persons. Tables 2 and 3 include households with heads below age 60, and Tables 4 and 5 include persons aged 25 59. PATHWAYS The and Inequality Report 2016
16 poverty TABLE 1. Overall Rates in High- and Middle-Income Countries, 2010 Income Level, 2010 A B C D E F 50% DHI U.S. Line ANGLOPHONE Australia H 29.1 13.8 15.3 25.8 5.7 20.1 Canada H 33.0 12.0 21.1 28.2 6.4 21.8 Ireland H 43.6 9.3 34.3 43.1 8.4 34.7 United Kingdom H 33.6 10.1 23.5 33.0 9.2 23.8 United States H 31.2 16.2 15.1 26.1 9.2 16.9 Group average 34.1 12.3 21.8 31.2 7.8 23.5 CONTINENTAL EUROPE France H 41.2 8.6 32.5 39.6 7.0 32.6 Germany H 40.9 11.3 29.6 39.5 8.2 31.3 Luxembourg H 31.4 5.7 25.7 22.7 1.5 21.2 Netherlands H 31.3 4.8 26.6 30.0 3.1 26.9 Group average 36.2 7.6 28.6 32.9 5.0 28.0 EASTERN EUROPE Czech Republic H 28.6 6.3 22.2 37.4 17.2 20.3 Estonia H 34.5 11.7 22.9 51.2 44.1 7.0 Georgia M 38.2 18.6 19.6 96.4 95.6 0.8 Hungary H 69.9 7.0 62.9 85.6 44.8 40.8 Poland H 36.9 9.2 27.7 59.0 40.0 19.0 Russia M 32.0 13.1 18.9 49.6 33.1 16.5 Serbia M 44.8 14.1 30.7 81.7 66.5 15.1 Slovak Republic H 29.2 7.3 21.8 40.7 22.0 18.7 Slovenia H 35.5 10.2 25.2 39.2 14.0 25.1 Group average 38.8 10.8 28.0 60.1 41.9 18.1 NORDIC EUROPE Denmark H 32.3 6.1 26.2 30.7 3.2 27.5 Finland H 32.4 7.2 25.2 30.7 4.0 26.7 Iceland H 22.9 5.7 17.2 19.9 3.3 16.6 Norway H 31.4 7.0 24.4 27.8 3.4 24.4 Group average 29.8 6.5 23.3 27.3 3.5 23.8 SOUTHERN EUROPE Greece H 37.5 13.0 24.4 43.1 22.2 20.9 Italy H 37.4 12.7 24.7 42.1 18.1 24.0 Spain H 38.7 14.7 24.0 42.2 20.3 22.0 Group average 37.9 13.5 24.4 42.5 20.2 22.3 LATIN AMERICA Brazil M 33.9 17.2 16.7 80.0 75.2 4.8 Colombia M 21.5 19.7 1.8 87.4 87.5-0.1 Mexico M 27.1 19.8 7.3 82.2 79.0 3.2 Panama M 28.6 22.1 6.5 73.1 70.6 2.5 Peru M 29.4 25.2 4.2 82.2 81.4 0.8 Uruguay M 33.5 15.2 18.3 74.3 65.9 8.4 Group average 29.0 19.9 9.1 79.9 76.6 3.3 OTHER Israel H 32.7 20.2 12.5 39.9 30.5 9.4 South Africa M 47.1 25.8 21.3 82.9 83.9-1.0 Taiwan H 12.3 11.7 0.6 11.7 10.9 0.8 Source: Luxembourg Income Study (LIS) Database, authors calculations. A table of poverty rates for all countries and years in the full LIS Database is available in an online appendix to this report. PATHWAYS The and Inequality Report 2016
poverty head 17 FIGURE 1. Overall Rates in High- and Middle-Income Countries Based on Income, 2010 Absolute Rate (Percent Below U.S. Line) 100 90 80 70 60 50 40 30 20 10 0 HU PL EE RS RU SK GR ES CZ IT SI TW IS DKNOFI IE UK CA FR DE AU NL LU UY U.S. 0 5 10 15 20 25 30 35 Relative Rate (Percent Below 50% of Median DHI) GE MX IL BR PA ZA PE CO Source: Luxembourg Income Study (LIS) Database, authors calculations. One of our core interests is the poverty-reducing effects of taxes and transfers. We omit older households (i.e., with heads aged 60+) and persons (i.e., age 60+) in order to lay aside income support programs that primarily serve retirees; those programs have distinct logics, and a full assessment of them is outside the scope of this report. Furthermore, in Tables 2 5, we examine a subset of the countries covered in Table 1. In these latter tables, we focus on 20 countries, selected because they are all high-income countries (and thus the most straightforward comparators for the U.S.) and because they all contain data on pre-tax market income. Having pre-tax data allows us to calculate comparable estimates of poverty reduction, capturing both taxes and transfers, across all 20 of these countries. Throughout this report, we group countries by regime type, drawing on a widely used social science framework that classifies countries according to social policy designs; these regimes, of course, overlap standard geographic groupings as well. We make use of these clusters however imperfect they are because they provide an organizing framework for assessing cross-national variation. They help us to identify empirical patterns across countries, and they bring into relief the importance of policy configurations for poverty reduction. The U.S. in Comparative Perspective We begin with an overview of poverty across nearly three dozen countries, deploying standard definitions of the poverty line 7 and of country income levels (see Table 1). 8 The core question here, it may be recalled, is whether the conventional view that the U.S. system produces high poverty rates is robust across a more expansive range of countries and with regard to multiple poverty definitions. Table 1 shows that, among high-income countries (marked H), U.S. poverty rates are indeed exceptionally high. When we consider disposable-income poverty in relative terms (see column B), which is the most common approach in comparative poverty studies, the U.S. poverty rate (16.2%) exceeds that reported in every other high-income country included here, with the sole exception of Israel (20.2%). The crossnational variation is substantial. In 12 of these high-income countries, poverty rates based on this measure are below 10 percent; in the Netherlands, the rate is below 5 percent. Is the U.S. just as exceptional among a broader range of countries? The simple answer: No. Among middle-income PATHWAYS The and Inequality Report 2016
18 poverty countries (marked M), the result is, in fact, quite different. Nearly all of the middle-income countries in this study report higher relative poverty rates than are seen in the U.S. Countries with greater relative poverty include five Latin American countries (Brazil, Colombia, Mexico, Panama, Peru), as well as Georgia and South Africa. Thus, the well-known exceptionalism of American 9 poverty pertains to rich countries; it is not a universal result. Again, the variation is not insubstantial; relative poverty in South Africa exceeds 25 percent. When we consider columns A, B, and C together, we understand more fully what drives the high rate of U.S. poverty vis-à-vis disposable income. First, U.S. disposable income poverty (16.2%) is the ninth highest among these 34 countries. Second, in contrast, U.S. market income poverty (31.2%) is 25th highest. Third, the level of poverty reduction (15.1 percentage points) is the seventh lowest. We can thus conclude: The high disposable income poverty rate in the U.S. is driven more by meager poverty reduction than by high market poverty rates. The only countries that reduce poverty, via taxes and transfers, less than the U.S. does are four Latin American countries, Israel, and Taiwan (and, in Taiwan, little poverty reduction is needed). What about poverty with respect to a fixed real income poverty line, often called absolute poverty? When we switch our framework to use the official U.S. poverty line as our anchor, our results shift again and markedly. Using that U.S. line, disposable income poverty in the U.S. is 9.2 percent (see column E). The absolute poverty rate is higher in all 10 middle-income countries (not surprisingly). It is also higher in 11 other high-income countries including all of the Eastern and Southern European countries in our analysis. At the same time, U.S. absolute poverty remains high among a core group of rich comparator countries (see Figure 1). 10 American poverty, using the U.S. line, exceeds that reported in all of the other Anglophone countries (except the UK), as well as in all of the Continental and Nordic cases. In cross-national terms, American poverty stands out and that is true for both relative poverty and absolute poverty. Children as a Special Case It is well known that the rate of poverty among children is high in the U.S. We replicate this well-known result here: Among the 20 rich countries included in Table 2, the U.S. reports the highest rate of disposable income poverty among children: 21.1 percent (see column B). In general, (relative) child poverty is most prevalent in the U.S. and in the Southern European countries; it is least prevalent in the Nordic countries. As was found in Table 1, when we shift to absolute poverty, the U.S. falls to a middling position. In absolute terms, children are more likely to be poor in the Eastern and Southern European countries, but they are also less likely to be poor, compared to American children, in 11 of these 20 study countries. Nine of those 11 countries all but Luxembourg and Norway are less affluent than the U.S. It is also often noted that U.S. children are at greater risk of poverty than is the American population as a whole. Indeed, vis-à-vis relative poverty, our results indicate that American children are, remarkably, 30 percent more likely to live in poverty than is the U.S. population overall (see Table 2, column H). Is this a widespread pattern? In fact, it is. In about threequarters of the 20 rich countries included here, children s poverty risk (vis-à-vis disposable income) is higher than that of all persons. This same result does not, however, hold for market income. In all five country groups, children s risk of living in poverty when we consider market income (see column G) is less than the risk reported in their countries more generally. Relative to overall poverty in their own countries, children are 93 percent as likely to be market-income poor (on average) in the Anglophone group, 65 68 percent in the Continental and Eastern European countries, and 56 percent in both the Nordic and Southern European clusters. That general pattern is not surprising; children live in households with adults whose main income source is the labor market (see Table 2), whereas country populations as a whole (see Table 1) include households with elderly adults who have left paid work for partial or full retirement (and who thus have little or no market income). When we shift to disposable income poverty (see column H), the story changes dramatically. As noted, in most of these study countries, children s poverty risk is higher than in their countries overall. It is lower in only four countries: three Nordic countries (Denmark, Finland, and Norway) and (by a smaller margin) in the UK. The final point to be made, as shown in column I, is that poverty reduction is much more limited for children than for persons overall (with the UK as an exception). While children are not expected to be as dependent on income augmentation by the state as are other demographic groups (such as older persons), state interventions still matter. Consider a comparison of the U.S. and the UK. The two have similar rates of market income poverty: 30.3 percent in the U.S. and 33.8 percent in the UK (see column A). But they are extremely different with respect to disposable income poverty: 21.1 percent in the U.S. compared to 9.4 percent in the UK (see column B). This country pairing tells us that, regarding child PATHWAYS The and Inequality Report 2016
poverty 19 TABLE 2. Child Rates in High-Income Countries, 2010 A B C D E F G H I 50% DHI U.S. Line Ratio of all children to all persons (Table 2 compared to Table 1) 50% DHI Income (DHI) (MI Less DHI) Income (DHI) (MI Less DHI) Income (DHI) (MI Less DHI) ANGLOPHONE Australia 27.8 14.4 13.4 22.8 7.0 15.8 0.96 1.04 0.88 Canada 25.6 14.3 11.2 19.1 6.9 12.2 0.77 1.20 0.53 Ireland 42.1 10.1 32.0 41.6 9.2 32.4 0.97 1.09 0.93 United Kingdom 33.8 9.4 24.4 32.8 8.1 24.8 1.01 0.93 1.04 United States 30.3 21.1 9.2 23.2 12.1 11.1 0.97 1.30 0.61 Group average 31.9 13.9 18.1 27.9 8.7 19.3 0.93 1.11 0.80 CONTINENTAL EUROPE France 29.1 11.5 17.7 26.8 8.8 18.0 0.71 1.33 0.54 Germany 29.2 19.1 10.1 27.2 16.4 10.8 0.71 1.69 0.34 Luxembourg 25.3 9.4 15.9 12.1 1.7 10.4 0.80 1.65 0.62 Netherlands 11.4 6.3 5.1 9.7 3.2 6.5 0.36 1.33 0.19 Group average 23.8 11.6 12.2 19.0 7.5 11.4 0.65 1.50 0.42 EASTERN EUROPE Czech Republic 18.3 10.6 7.7 31.5 22.9 8.6 0.64 1.67 0.35 Estonia 22.4 12.7 9.7 43.3 36.9 6.4 0.65 1.09 0.42 Poland 25.4 12.0 13.4 54.3 46.8 7.5 0.69 1.30 0.48 Slovak Republic 21.2 13.2 8.0 38.9 31.1 7.7 0.73 1.81 0.37 Group average 21.8 12.1 9.7 42.0 34.4 7.6 0.68 1.47 0.41 NORDIC EUROPE Denmark 14.6 4.5 10.1 12.9 2.3 10.6 0.45 0.74 0.39 Finland 16.0 3.7 12.3 14.0 1.9 12.0 0.49 0.51 0.49 Iceland 17.2 7.4 9.8 13.3 3.5 9.8 0.75 1.30 0.57 Norway 16.5 5.2 11.3 12.1 1.8 10.4 0.53 0.74 0.46 Group average 16.1 5.2 10.9 13.1 2.4 10.7 0.56 0.82 0.48 SOUTHERN EUROPE Greece 17.9 17.3 0.6 25.6 27.4-1.8 0.48 1.33 0.02 Italy 21.2 19.1 2.1 27.5 25.2 2.3 0.57 1.50 0.08 Spain 24.9 20.6 4.3 30.0 27.0 3.0 0.64 1.40 0.18 Group average 21.3 19.0 2.3 27.7 26.5 1.1 0.56 1.41 0.10 Source: Luxembourg Income Study (LIS) Database, authors calculations. Child = under age 18 living with household head under age 60. poverty, policy matters, and it matters a lot. Type Matters Another stylized fact, widely reported in the U.S., is that persons in households headed by single mothers are especially likely to be poor. In Table 3, we see that Americans living in single mother headed households are indeed at high risk of (disposable income) poverty fully 36.5 percent are poor (see column B). That is the highest rate among these 20 countries. What is driving this high poverty rate in the U.S.? Consider the market-income poverty results (see column A). Note that there are three countries, among these 20, where marketincome poverty among single mother headed households is similar or higher than it is in the U.S. (where it is 50.1%): Ireland (70.9%), the UK (56.0%), and Luxembourg (50.6%). Yet in these three comparison cases, disposable-income poverty is substantially lower than it is in the U.S. (where it is 36.5%): Ireland (22.7%), the UK (12.3%), and Luxembourg (25.9%). These results indicate that that the impact of tax-and-transfer policy varies across national contexts, and it varies extensively. Our results reveal that Americans in two-parent households PATHWAYS The and Inequality Report 2016
20 poverty TABLE 3. Rates by Structure, 2010 A B C D E F G H I Single-Mother s / Single-Mother s Two-Parent s Two-Parent s ANGLOPHONE Australia 46.5 22.5 24.0 13.1 6.9 6.2 3.5 3.3 3.9 Canada 40.8 26.0 14.8 14.9 7.6 7.3 2.7 3.4 2.0 Ireland 70.9 22.7 48.2 27.3 5.5 21.8 2.6 4.1 2.2 United Kingdom 56.0 12.3 43.7 17.0 6.7 10.2 3.3 1.8 4.3 United States 50.1 36.5 13.7 15.5 9.7 5.8 3.2 3.8 2.4 Group average 52.9 24.0 28.9 17.5 7.3 10.3 3.1 3.3 2.9 CONTINENTAL EUROPE France 45.4 20.9 24.5 18.2 7.0 11.2 2.5 3.0 2.2 Germany 47.8 28.0 19.8 10.9 5.3 5.6 4.4 5.3 3.5 Luxembourg 50.6 25.9 24.8 16.0 5.2 10.8 3.2 5.0 2.3 Netherlands 38.2 14.5 23.8 4.9 2.2 2.7 7.8 6.6 8.8 Group average 45.5 22.3 23.2 12.5 4.9 7.6 4.5 5.0 4.2 EASTERN EUROPE Czech Republic 33.5 22.1 11.4 9.7 5.3 4.4 3.4 4.1 2.6 Estonia 28.7 23.6 5.1 13.9 8.6 5.3 2.1 2.8 1.0 Poland 36.9 15.6 21.3 18.7 9.6 9.1 2.0 1.6 2.3 Slovak Republic 22.0 13.5 8.5 10.5 7.3 3.2 2.1 1.8 2.7 Group average 30.3 18.7 11.6 13.2 7.7 5.5 2.4 2.6 2.1 NORDIC EUROPE Denmark 31.7 6.8 24.9 6.7 2.1 4.5 4.8 3.2 5.5 Finland 35.3 10.3 25.0 11.4 2.5 8.9 3.1 4.2 2.8 Iceland 37.5 16.0 21.5 6.8 2.8 4.1 5.5 5.8 5.3 Norway 37.5 9.4 28.1 7.0 1.5 5.4 5.4 6.2 5.2 Group average 35.5 10.6 24.9 8.0 2.2 5.7 4.7 4.8 4.7 SOUTHERN EUROPE Greece 32.2 22.6 9.6 16.0 14.1 1.9 2.0 1.6 5.1 Italy 36.1 25.1 11.0 18.0 15.3 2.7 2.0 1.6 4.1 Spain 38.0 28.4 9.6 20.4 15.5 5.0 1.9 1.8 1.9 Group average 35.4 25.3 10.1 18.1 15.0 3.2 2.0 1.7 3.7 Source: Luxembourg Income Study (LIS) Database, authors calculations. Universe restricted to households with heads below age 60. are also at comparatively high risk of poverty (see column E) although the U.S. outcome (9.7%) is not as exceptional as it is for single mother headed households. Persons in twoparent households in all three Southern European countries are more likely to be poor than are their U.S. counterparts. Overall, while the magnitudes vary across countries, we do see a somewhat universal pattern. In all 20 countries, persons in households headed by a single mother are more likely to be poor both before and after taxes and transfers than are persons in households headed by two parents. It is also the case that poverty reduction is always greater in single-parent households but not extensive enough to equalize poverty rates at the end of the day (that is, disposable income poverty) between the two household types. Education Matters One s economic prospects, including the risk of being poor, are, of course, shaped by another crucial demographic factor: one s own educational attainment. In Table 4, we report poverty rates for persons (aged 25 59) with low, medium, and high levels of educational attainment. Education systems are complex, so measuring educational attainment with perfect institutional comparability across countries is not possible. Nevertheless, nearly everywhere, PATHWAYS The and Inequality Report 2016
poverty 21 TABLE 4. Rates by Educational Attainment among Persons Aged 25 59, 2010 A B C D E F G H I Low Education Medium Education High Education ANGLOPHONE Australia 30.1 15.0 15.1 17.3 9.4 7.9 10.0 5.9 4.1 Canada 38.2 20.6 17.6 25.1 13.8 11.3 16.1 9.3 6.8 Ireland 55.5 13.1 42.4 35.1 7.7 27.4 16.3 5.1 11.2 United Kingdom 49.2 15.5 33.7 20.9 7.6 13.3 9.5 5.8 3.7 United States 51.3 37.8 13.5 25.4 15.9 9.5 9.7 5.9 3.8 Group average 44.9 20.4 24.5 24.8 10.9 13.9 12.3 6.4 5.9 CONTINENTAL EUROPE France 40.9 13.6 27.2 21.5 7.5 14.1 10.3 4.6 5.8 Germany 35.8 19.7 16.2 19.0 11.5 7.5 10.6 6.7 3.9 Luxembourg 31.6 11.2 20.4 15.8 4.7 11.1 6.2 2.8 3.4 Netherlands 22.0 5.2 16.8 11.9 4.2 7.7 7.7 3.6 4.1 Group average 32.6 12.4 20.1 17.1 7.0 10.1 8.7 4.4 4.3 EASTERN EUROPE Czech Republic 44.5 19.1 25.4 14.0 5.8 8.2 4.7 1.8 2.9 Estonia 40.7 24.9 15.8 23.9 15.0 8.9 7.0 4.3 2.8 Poland 54.5 24.3 30.1 26.9 9.5 17.5 7.7 2.0 5.8 Slovak Republic 52.2 26.7 25.5 16.7 7.9 8.8 6.0 3.3 2.7 Group average 48.0 23.7 24.2 20.4 9.5 10.8 6.3 2.8 3.5 NORDIC EUROPE Denmark 32.0 6.2 25.8 12.3 3.9 8.4 8.1 3.6 4.6 Finland 33.5 11.1 22.4 20.8 7.7 13.1 7.1 2.8 4.3 Iceland 21.0 5.3 15.7 14.8 7.6 7.2 7.4 4.4 3.0 Norway 31.7 6.7 25.0 13.4 3.3 10.1 8.4 4.2 4.2 Group average 29.6 7.3 22.2 15.3 5.6 9.7 7.8 3.8 4.0 SOUTHERN EUROPE Greece 33.7 21.8 11.9 21.3 13.3 8.0 12.3 4.8 7.5 Italy 33.8 20.9 12.9 15.3 7.9 7.4 7.9 3.5 4.4 Spain 36.6 21.0 15.6 19.7 11.7 8.0 10.8 6.3 4.6 Group average 34.7 21.2 13.5 18.8 10.9 7.8 10.4 4.9 5.5 Source: Luxembourg Income Study (LIS) Database, authors calculations. poverty rates based on both market and disposable income are highest in the least educated group, lower in the medium-educated group, and lower yet in the most highly educated group. 11 Not surprisingly, then, the reverse holds for poverty reduction; in all 20 study countries, it falls as educational attainment rises. The U.S. result, while not entirely exceptional, is notable. Among those with the lowest attainment (see column B), Americans are the most likely to be poor (37.8%); this reflects the general pattern of high poverty in the U.S., as shown throughout this report. Among those with the highest attainment (see column H), Americans are the fourth most likely to be poor (5.9%; tied with Australia); highly educated persons are more likely to be poor in Spain (6.3%), Germany (6.7%), and Canada (9.3%). In the U.S., those with the least education (as a group) are more than six times as likely to be poor as those with the most education (results not shown). That ratio 6.4 is exceeded in only three countries (all in Eastern Europe). In the U.S., as everywhere, education matters, and again, it matters a lot. Paid Work Matters In the U.S., we take it for granted that working for pay and especially commanding high earnings is a poverty prevention tool. Is that reliably the case in the U.S.? And elsewhere? And for whom? In Table 5, we report disposable income pov- PATHWAYS The and Inequality Report 2016
22 poverty TABLE 5. Rates by Gender and Level of among Persons Aged 25 59, 2010 A B C D E F G H I J K L Gender Gap All persons 25 59 Men 25 59 Women 25 59 (Women minus Men) No Low Medium- High No Low Medium- High No Low Medium- High No Low Medium- High ANGLOPHONE Australia 31.0 8.6 0.3 40.0 8.5 0.1 26.4 8.6 0.5-13.6 0.0 0.4 Canada 37.2 20.6 0.3 42.0 20.6 0.2 34.1 20.6 0.5-7.9 0.0 0.3 Ireland 18.6 7.7 0.0 20.9 9.1 0.0 17.0 6.1 0.1-3.9-3.0 0.1 United Kingdom 24.1 9.1 0.5 32.2 9.5 0.3 19.0 8.7 0.7-13.2-0.8 0.4 United States 37.9 21.3 0.5 42.0 20.6 0.2 35.6 22.1 0.8-6.5 1.5 0.6 Group average 29.7 13.4 0.3 35.4 13.7 0.2 26.4 13.2 0.5-9.0-0.4 0.3 CONTINENTAL EUROPE France 28.4 12.0 0.4 32.6 12.4 0.2 26.0 11.6 0.5-6.6-0.8 0.3 Germany 33.0 15.4 3.7 50.6 12.7 2.4 25.7 17.8 4.9-24.9 5.1 2.6 Luxembourg 13.1 13.1 0.7 14.9 12.6 0.2 12.3 13.8 1.3-2.6 1.2 1.0 Netherlands 12.4 7.7 0.5 20.1 8.0 0.0 9.0 7.3 0.9-11.1-0.7 0.9 Group average 21.7 12.1 1.3 29.6 11.4 0.7 18.3 12.6 1.9-11.3 1.2 1.2 EASTERN EUROPE Czech Republic 18.4 9.3 0.2 23.3 8.9 0.1 16.7 9.7 0.2-6.6 0.8 0.0 Estonia 36.0 22.2 0.5 45.4 22.9 0.0 28.7 21.5 1.1-16.7-1.4 1.1 Poland 19.6 9.6 0.7 24.3 11.2 0.6 16.9 7.7 0.7-7.4-3.6 0.1 Slovak Republic 24.6 11.1 0.8 27.8 11.0 0.8 22.4 11.1 0.8-5.4 0.1 0.0 Group average 24.6 13.0 0.5 30.2 13.5 0.4 21.2 12.5 0.7-9.0-1.0 0.3 NORDIC EUROPE Denmark 16.2 7.1 0.2 21.2 7.1 0.0 11.9 7.1 0.3-9.2 0.0 0.2 Finland 27.4 9.1 0.0 33.6 10.3 0.0 21.8 7.9 0.0-11.8-2.4 0.0 Iceland 17.0 15.0 0.1 22.9 13.6 0.0 13.5 16.5 0.3-9.4 2.9 0.3 Norway 16.2 8.9 0.2 20.2 9.5 0.1 13.1 8.2 0.3-7.2-1.3 0.2 Group average 19.2 10.0 0.1 24.5 10.1 0.0 15.1 9.9 0.2-9.4-0.2 0.2 SOUTHERN EUROPE Greece 25.0 19.9 1.4 29.6 23.5 1.1 22.6 15.0 1.7-7.0-8.5 0.7 Italy 26.5 19.4 0.5 34.8 24.2 0.7 23.7 13.6 0.2-11.1-10.5-0.5 Spain 32.6 18.4 1.3 37.9 19.5 1.1 29.2 17.1 1.5-8.7-2.4 0.4 Group average 28.0 19.3 1.0 34.1 22.4 1.0 25.2 15.3 1.2-8.9-7.1 0.2 Source: Luxembourg Income Study (LIS) Database, authors calculations. erty rates for those with zero earnings, low earnings, and medium-high earnings. We do so for all persons aged 25-59 and for men and women separately. As with educational attainment, we find some universal patterns: everywhere, for men and women alike, the risk of poverty falls as one s own attachment to the labor market increases (from none, to low, to medium-high) and the gradient is steep across all country clusters. Those with no earnings experience high rates of living in disposable income poverty on average, 29.7 percent in the Anglophone countries, 21.7 percent in the Continental European countries, 24.6 percent in the Eastern European countries, 19.2 percent in the Nordic countries, and 28.0 percent in the Southern European countries (see column A). In contrast, medium-high earners face much lower risks of poverty everywhere, typically poverty rates of 1 percent or less (see column C). The final panel reveals that gender matters especially among those who are potentially more vulnerable to poverty. Note that, in all 20 countries, among persons with no earnings, men are substantially more likely to be poor than are their female counterparts (see column J). That finding reflects persistent gendered divisions in paid work; women with no earnings are more likely to be partnered with (and sharing household income with) earners than are men. Women with PATHWAYS The and Inequality Report 2016
poverty 23 no earnings are also more likely to have co-resident children, so they are eligible for more income transfers conditioned on the presence of children. In the U.S., those with no earnings (as a group) are substantially more likely to be poor (37.9%) than are those with earnings in the bottom third of the distribution (21.3%). Nonearners are vastly more likely to live in poverty than are high earners (0.5%). That is a general pattern across this group of rich countries. Conclusion In this report, we find that, using a relative poverty standard and considering disposable household income, the U.S. poverty rate exceeds that reported in all of the other high-income countries in our study, with the sole exception of Israel. At the same time, the well-known exceptionalism of American relative poverty is not a universal result; most of the middleincome countries in this study report higher relative poverty rates than are seen in the U.S. Our results help us to understand more fully what drives the comparatively high rate of U.S. relative (disposable income) poverty. The high rate in the U.S. is driven more by meager poverty reduction than by high market poverty rates. The only countries that reduce poverty, via taxes and transfers, less than the U.S. does are four Latin American countries, Israel, and Taiwan. When we use the official U.S. poverty line the absolute line as our anchor, our results shift. The absolute poverty rate (vis-à-vis disposable income) is, not surprisingly, higher than it is in the U.S. in all 10 middle-income countries included here. Absolute poverty is also higher in 11 other high-income countries in our study including all of the Eastern and Southern European countries. What is more surprising is that absolute poverty in the U.S. is higher than in a core group of rich comparator countries. American poverty, using the U.S. line, exceeds that reported in nearly all of the other Anglophone countries, as well as in all of the Continental and Nordic cases. All told, in cross-national terms, especially compared to other rich countries, American poverty stands out and that is true for both relative poverty and absolute poverty. We also find that child poverty is especially high in the U.S. Using the most common indicator, relative disposable income poverty, among the 20 rich countries included here, the U.S. reports the highest percentage of children living in poverty. One key finding is that U.S. tax-and-transfer policies reduce child poverty less than do policies in many other countries. Likewise, we find that Americans living in single mother headed households face the highest poverty rate among our 20 study countries. Americans in two-parent households also face a comparatively high risk of poverty although the U.S. outcome is not as exceptional as it is for single mother headed households. In addition, we find two general patterns across these study countries: the risk of poverty falls as educational attainment rises and as labor market attachment increases. In the case of education, the gradient in the U.S. is comparatively steep (i.e., education is an especially important poverty prevention tool); with respect to labor market attachment, the U.S. gradient is not unusual. Finally, our results indicate that national-level policies and institutions play a major role in shaping poverty outcomes. Returning to our first table, and considering relative (disposable income) poverty, we see that, on average, the Anglophone countries produce more poverty (12.3%), than do the Eastern European (10.8%), Continental (7.6%), and Nordic countries (6.5%); they produce less relative poverty than reported in the Southern European (13.5%) and Latin American (19.9%) countries studied here. To close, we conclude that American policy makers should look abroad for lessons about poverty and poverty reduction lessons that are both inspirational and cautionary. Janet C. Gornick is Professor of Political Science and Sociology at the Graduate Center, City University of New York, and Director of LIS, the cross-national data center in Luxembourg. Markus Jäntti is Professor of Economics at the Institute for Social Research, Stockholm University. PATHWAYS The and Inequality Report 2016
24 poverty NOTES 1. For results based on the LIS data, see, e.g., Brady, 2009; Gornick and Jäntti, 2012, 2010, 2009; Rainwater and Smeeding, 2003; also see the LIS Inequality and Key Figures: http://www.lisdatacenter.org/data-access/keyfigures/inequality-and-poverty/. See also OECD, 2015, 2011, 2008, and Eurostat, 2015. 2. We define market income (MI) as pre-taxpre-transfer income, and disposable household income (DHI) as post-tax-post-transfer income. For convenience, we shorthand DHI as disposable income. MI includes income from labor, from selected sources of capital, and from private transfers. DHI adjusts market income by subtracting direct taxes paid out (i.e., income taxes and social contributions) and by adding the value of public transfers received. (Counted income is not reduced to account for non-discretionary expenditures other than direct taxes.) Income in all households is adjusted for family size, using the widely-used square root equivalence scale. 3. We use a tripartite education classification in our analyses. Low educational attainment includes those who have not completed upper secondary education; medium refers to those who have completed upper secondary education or non-specialized vocational education; and high includes those who have completed post-secondary education, specialized vocational education, and beyond. LIS provides standard recodes for most countries, based on an international classification system. Where LIS did not provide recodes, we constructed them, adhering to these educational cutoffs as closely as possible. 4. We define low earners as those whose annual earnings fall in the bottom third of the earnings distribution (among those with positive earnings) and medium-high earners as those in the top two-thirds. These distributions are countryspecific and gender-specific. 5. U.S. results cited in this introduction are reported in Tables 2 5 in this report. 6. See www.lisdatacenter.org for a detailed description of the Luxembourg Income Study (LIS) Database. The LIS Database contains approximately 300 datasets from nearly 50 countries. The data are available in repeated cross sections (1980, 1985, 1990, 1995, 2000, 2004, 2007, 2010); as of this writing, LIS is nearing completion of the 2010 wave and has started making available datasets from 2013. 7. In Tables 1 and 2, we use two different poverty lines. The first one, the relative line, is drawn at 50 percent of median DHI (where the DHI-based line is used to calculate both market income poverty and disposable income poverty). The 50%-of-median line is countryspecific, meaning that relative poverty refers to income relative to others in the same country. The second one, the absolute line, is set at the level of the U.S. poverty line, which is converted to international dollars, adjusted for purchasing power parities (PPPs). All results in the rest of the tables pertain to relative poverty. We define poverty reduction using a simple accounting framework: it is the MI-based poverty rate minus the DHI-based poverty rate. 8. The World Bank classifies all the world s countries as high income, upper-middle income, lower-middle income, and low income. We use the terms high-income countries and rich countries interchangeably. We use the term middle-income to refer to upperand lower-middle income countries. 9. For convenience, we use the term American as an adjective referring to U.S. residents or conditions in the U.S. We understand that not all U.S. residents are American by nationality, and that the U.S. is not alone in the Americas. 10. In Figure 1, the abbreviations are as follows: Australia (AU), Brazil (BR), Canada (CA), Colombia (CO), Czech Republic (CZ), Denmark (DK), Estonia (EE), Finland (FI), France (FR), Georgia (GE), Germany (DE), Greece (GR), Hungary (HU), Iceland (IS), India (IN), Ireland (IE), Israel (IL), Italy (IT), Luxembourg (LU), Mexico (MX), Netherlands (NL), Norway (NO), Panama (PA), Peru (PE), Poland (PL), Russia (RU), Serbia (RS), Slovak Republic (SK), Slovenia (SI), South Africa (ZA), Spain (ES), Taiwan (TW), United Kingdom (UK), United States (U.S.). 11. Iceland and Norway are minor exceptions to the pattern (with respect to disposable income poverty). ADDITIONAL RESOURCES Brady, David. 2009. Rich Democracies, Poor People: How Politics Explain. New York: Oxford University Press. Gornick, Janet C., and Markus Jäntti. 2012. Child in Cross-National Perspective: Lessons from the Luxembourg Income Study. Children and Youth Services Review 34: 558 568. Gornick, Janet C., and Markus Jäntti. 2010. Women,, and Social Policy Regimes: A Cross-National Analysis. In Peter Saunders and Roy Sainsbury (eds), Social Security, and Social Exclusion in Rich and Poorer Countries. International Studies on Social Security - Volume 16, Antwerp: Intersentia, 63 95. Gornick, Janet C., and Markus Jäntti. 2009. Child in Upper-Income Countries: Lessons from the Luxembourg Income Study. In Sheila B. Kamerman, Shelley Phipps and Asher Ben-Arieh (eds), From Child Welfare to Child Wellbeing: An International Perspective on Knowledge in the Service of Making Policy. New York: Springer Publishing Company, 339 368. Rainwater, Lee, and Timothy M. Smeeding. 2003. Poor Kids in a Rich Country. New York: Russell Sage Foundation. Organization for Economic Cooperation and Development (OECD). 2015. In It Together: Why Less Inequality Benefits All. Paris: OECD. OECD. 2011. Divided We Stand: Why Inequality Keeps Rising. Paris: OECD. OECD. 2008. Growing Unequal? Income Distribution and in OECD Countries. Paris: OECD. Eurostat. 2015. The EU in the World 2015 Edition. Luxembourg: Publications Office of the European Union. PATHWAYS The and Inequality Report 2016